2:20 p.m.
Hi,
I'm a bit confused about the situation with the package FEP (current
version). We specified no-dataset-ok and the info window of the package
(1.3) says "Data requirement: none". However, when I "execute" the
package from the listing window I get an error window "Please open data
file first".
This does not happen with other packages that don't need a dataset.
However, FEP specifies a GUI main function, and the other packages don't
seem to do that (e.g. extra.gfn, corresp.gfn), so maybe that's the
difference?
thanks,
sven
2:30 p.m.
On Sun, 10 Dec 2017, Sven Schreiber wrote:
Hi,
I'm a bit confused about the situation with the package FEP (current
version). We specified no-dataset-ok and the info window of the package (1.3)
says "Data requirement: none". However, when I "execute" the package
from the
listing window I get an error window "Please open data file first".
This does not happen with other packages that don't need a dataset.
However, FEP specifies a GUI main function, and the other packages don't seem
to do that (e.g. extra.gfn, corresp.gfn), so maybe that's the difference?
Hmm, not sure but I'll look into it. Maybe has something to do with
the fact that the FEP package includes a data file? (Though that
shouldn't in itself have the effect described.)
Allin
3:26 p.m.
Am 10.12.2017 um 20:30 schrieb Allin Cottrell:
On Sun, 10 Dec 2017, Sven Schreiber wrote:
> Hi,
>
> I'm a bit confused about the situation with the package FEP (current
> version). We specified no-dataset-ok and the info window of the
> package (1.3) says "Data requirement: none". However, when I
"execute"
> the package from the listing window I get an error window "Please open
> data file first".
>
> This does not happen with other packages that don't need a dataset.
> However, FEP specifies a GUI main function, and the other packages
> don't seem to do that (e.g. extra.gfn, corresp.gfn), so maybe that's
> the difference?
Hmm, not sure but I'll look into it. Maybe has something to do with the
fact that the FEP package includes a data file? (Though that shouldn't
in itself have the effect described.)
It looks as if upgrading to the latest snapshot solved the problem! (I
had a slightly older snapshot before.)
thanks,
sven
3:35 p.m.
Am 10.12.2017 um 21:26 schrieb Sven Schreiber:
Am 10.12.2017 um 20:30 schrieb Allin Cottrell:
> Hmm, not sure but I'll look into it. Maybe has something to
do with
> the fact that the FEP package includes a data file? (Though that
> shouldn't in itself have the effect described.)
It looks as if upgrading to the latest snapshot solved the problem! (I
had a slightly older snapshot before.)
Whoops, no -- sorry for the noise. I ignored that this time I did
already have a dataset loaded... So the problem is still there.
-s
4:14 p.m.
On Sun, 10 Dec 2017, Sven Schreiber wrote:
Am 10.12.2017 um 21:26 schrieb Sven Schreiber:
> Am 10.12.2017 um 20:30 schrieb Allin Cottrell:
>> Hmm, not sure but I'll look into it. Maybe has something to do with the
>> fact that the FEP package includes a data file? (Though that shouldn't in
>> itself have the effect described.)
>
> It looks as if upgrading to the latest snapshot solved the problem! (I had
> a slightly older snapshot before.)
Whoops, no -- sorry for the noise. I ignored that this time I did already
have a dataset loaded... So the problem is still there.
I found it: nothing to do with packing a data file -- the point is
that although FEP nominally does not require a dataset in place its
gui-main function requires a series argument. Which tends to imply
that its no-data-required status only really applies to use via
script/command-line. This is a case we haven't seen before. Maybe
there's a way around it...
Allin
6:01 p.m.
Am 10.12.2017 um 22:14 schrieb Allin Cottrell:
I found it: nothing to do with packing a data file -- the point is
that
although FEP nominally does not require a dataset in place its gui-main
function requires a series argument. Which tends to imply that its
no-data-required status only really applies to use via
script/command-line. This is a case we haven't seen before. Maybe
there's a way around it...
Right, but that basically means that gretl is right in demanding a
dataset when the GUI-main function is going to be called.
Maybe then nothing needs to be changed after all...?
thanks,
sven
9:58 p.m.
On Mon, 11 Dec 2017, Sven Schreiber wrote:
Am 10.12.2017 um 22:14 schrieb Allin Cottrell:
> I found it: nothing to do with packing a data file -- the point is that
> although FEP nominally does not require a dataset in place its gui-main
> function requires a series argument. Which tends to imply that its
> no-data-required status only really applies to use via script/command-line.
> This is a case we haven't seen before. Maybe there's a way around it...
Right, but that basically means that gretl is right in demanding a dataset
when the GUI-main function is going to be called.
Maybe then nothing needs to be changed after all...?
Maybe, but I think we should probably offer to open the package's
sample script, which is what we do for packages that explicitly
require a dataset, when there's no dataset loaded.
Allin
3:42 a.m.
Am 11.12.2017 um 03:58 schrieb Allin Cottrell:
On Mon, 11 Dec 2017, Sven Schreiber wrote:
> Am 10.12.2017 um 22:14 schrieb Allin Cottrell:
>
>> I found it: nothing to do with packing a data file -- the point is
>> that although FEP nominally does not require a dataset in place its
>> gui-main function requires a series argument. Which tends to imply
>> that its no-data-required status only really applies to use via
>> script/command-line. This is a case we haven't seen before. Maybe
>> there's a way around it...
>
> Right, but that basically means that gretl is right in demanding a
> dataset when the GUI-main function is going to be called.
> Maybe then nothing needs to be changed after all...?
Maybe, but I think we should probably offer to open the package's sample
script, which is what we do for packages that explicitly require a
dataset, when there's no dataset loaded.
Yes, that sounds good.
-sven
1:01 p.m.
Am 11.12.2017 um 03:58 schrieb Allin Cottrell:
On Mon, 11 Dec 2017, Sven Schreiber wrote:
> Am 10.12.2017 um 22:14 schrieb Allin Cottrell:
>
>> I found it: nothing to do with packing a data file -- the point is
>> that although FEP nominally does not require a dataset in place its
>> gui-main function requires a series argument. Which tends to imply
>> that its no-data-required status only really applies to use via
>> script/command-line. This is a case we haven't seen before. Maybe
>> there's a way around it...
>
> Right, but that basically means that gretl is right in demanding a
> dataset when the GUI-main function is going to be called.
> Maybe then nothing needs to be changed after all...?
Maybe, but I think we should probably offer to open the package's sample
script, which is what we do for packages that explicitly require a
dataset, when there's no dataset loaded.
Has this proposed solution/workaround been applied or just forgotten? I
saw the problem again in a Feb 5 snapshot.
thanks,
sven
2475
days inactive
2534
days old
8 comments
2 participants
participants (2)
-
Allin Cottrell -
Sven Schreiber